Uguqulelo lwe-IsiJamani olukhululekileyo
Uguqulelo lwe-IsiJamani lwesandi nevidiyo kumbhalo nge-AI. Ikhawulezayo, ichanekileyo, kwaye ikhululekile.
Indlela esebenza ngayo
- Yiya kwi- Free.ai Umshicileli
- Layisha phezulu ifayili yakho yevidiyo okanye yesandi ye-IsiJamani
- I-AI yethu ifumanisa ngokuzenzekelayo i-IsiJamani kwaye ibhalela
- Layisha ezantsi i-transcript yakho njengombhalo okanye i-SRT subtitle
IsiJamani Iimpawu Zokushicilela
- ✓Isebenza nge faster-whisper (i-MIT licensed)
- ✓Ubhaqo oluzenzekelayo lweelwimi lweIsiJamani
- ✓Inkxaso ye MP3, WAV, MP4, M4A, FLAC, kunye nezinye
- ✓Ii-timestamps kunye norhwebo lwangaphandle lwesihloko esingaphantsi (SRT)
- ✓Akukho mda wobungakanani befayili kwinkqubo ehlawulweyo
- ✓Iifayile ezifihlakeleyo
Iinkcukacha Zolwimi
| Iilwimi | IsiJamani |
| Ikhowudi ye ISO | de |
| Imodeli ye-AI | _Iifayile: |
| Ixabiso | Iinketho ze-GNU/Linux |
Iilwimi ezininzi
Bonisa Zonke IilwimiImibuzo ebuzwa rhoqo
I-Whisper enkulu-v3-i-turbo iwela kwinqanaba eliphezulu lempumelelo kwi IsiJamani - ngaphantsi kwe 7% yexabiso lemposiso yegama kwiimpawu eziqhelekileyo ze-benchmark. Kwimisebenzi ethetha ukuba i-audio yestudio ecocekileyo ibuyela kwi-perfect, kwaye i-audio yencoko isetyenziswa ngococeko oluncinci.(Inqanaba A, under 7% word error rate kwiseti ye-benchmark - sipapasha inqanaba elilungileyo le-WER kunokuba siqinisekise ukuthengisa.)
Ewe — IsiJamani ukudlulisa kuqala kusuka kwi-token pool yakho yosuku olusimahla. Isandi sibiza malunga ne-token 50 ngomzuzu, ngoko ke i-token pool yosuku olungaziwayo igubungela iiyure ezimbalwa zesandi ngosuku. Ii-akhawunti ezibhalisiweyo zifumana i-token pool enkulu ye-30,000-token ngosuku. Emva koko, ukudlulisa kuhlawulwe-njengoko-uhamba, nge-token top-ups ukusuka kwi-$1.
IsiJamani izicatshulwa ezibhaliweyo zibuyiselwa kwi UTF- 8 eqhelekileyo eneelwimi zopelo oluqhelekileyo.
MP3, WAV, M4A, FLAC, OGG, OPUS, ne WEBM zivunyelwa ngokuthe ngqo. Kwividiyo (MP4, MOV, MKV) sikhupha umculo osuka kumncedisi-ecaleni phambi kokuba siyithumelele kwi Whisper - awunakutshintsha nantoni na ngokwakho. Inkqubo efanayo yokuhambisa ngaphandle kolwazi lomntu ovela kulo, kubandakanya i IsiJamani.
Ukhuphelo olungaziwayo luya kufikelela kwi-500 MB kwifayili nganye. Ii-akhawunti ezibhalisiweyo ziye kwi-2 GB. Ukuphela kwexesha alikho umda onzima - iifayile ezide ziqhutywa ngokuzenzekelayo (iifestile zemizuzu engama-30 ezinamathele) kwaye zidityaniswe kwakhona kwi-transcript epheleleyo enee-timestamps eziqhubekayo. Iiyure ezininzi IsiJamani zokukhuphela (ipodcasts, izifundo ezipheleleyo, iintlanganiso) zisebenza kakuhle.
Ewe - ushicilelo lomculi luyasebenza ngokungagqibekanga kulo lonke ushicilelo lwe IsiJamani. Imveliso ihlukaniswe njenge Umculi 1 / Umculi 2 / Umculi 3 ngexesha elimiselweyo, ngoko udliwanondlebe, uxoxo lwepaneli, kunye neentlanganiso zeqela elininzi zibuyela umva zilabelwe. Ushicilelo luyasebenza kwimodeli eyahlukileyo kwaye lusebenza ngokufanayo kuzo zonke iilwimi esixhasayo.
Ewe - cofa i URL kwi /transcribe/youtube/ yeYouTube okanye /transcribe/podcast/ yepodcast feeds (Apple, Spotify, RSS). Sikhuphela umsindo, siqhuba nge Whisper nge ulwimi=de, kwaye sibuyisela uguqulelo ngee-timestamps kunye neelabels zomthumeli. Izinto eziqhelekileyo ze IsiJamani: iipodcast, izifundo, udliwanondlebe, kunye nemixholo yeYouTube ethe tyaba kwiIsiJamani zizinto eziqhelekileyo.
Whisper ibiza malunga ne-50 tokens ngomzuzu wesandi, ngoko urekhodo lweyure enye luyi ~3,000 tokens. Abaninzi abasebenzisi abachithanga nto - i-pool yemihla ngemihla ekhululekileyo ye-30,000 tokens iquka iiclip ezimfutshane, ii-voice notes, kunye ne-one-off podcasts. Ngapha koko, ukudlulisa kuhlawulwe-njengoko-uhamba, nge-token top-ups ukusuka kwi $1.
Ewe - zombini inqanaba lecandelo (imini nganye ~10-30 imizuzwana) kunye nexesha legama eliphakamileyo lifumaneka. Igama-lomgangatho lumiselwe kwi-VTT/SRT imveliso yesicatshulwa ukuze izihloko zidityaniswe umgca-nge-umgca. Kwi-API misela i-timestamps="igama" kwisiqu sesicelo. IsiJamani izicatshulwa ezibhaliweyo zibuyiselwa kwi UTF- 8 eqhelekileyo eneelwimi zopelo oluqhelekileyo.
Ewe. UTHANDA umsindo (inxalenye eninzi/ifomu-data, igama lendawo "ifayile") kwi /v1/transcribe / nge-language=de - okanye ushiye iparameter ye-language ukuze i-Whisper ifumane ngokuzenzekelayo. Ibuyisela i-JSON nge-transcript, ii-segments, i-timestamps, kunye nee-labels zomthumeli. Ubhekiso olupheleleyo kunye ne-SDK snippets kwi /api/.
Ewe - xa uguqulelo lugqityiwe, nqakraza Gcina okanye uncamathisele umbhalo kwi /translate/. IsiJamani idibanisa nezinye iilwimi zonke esizixhasayo (200+). Kwiimini zengxoxo uguqulelo luya ku /summarize/; xa kuthelekiswa, thumela ku /voice/tts/ ukuvelisa isandi kwiilwimi ezilindelweyo.
Whisper iqeqeshwe kwi 680K iiyure zesandi esingenasandi sehlabathi elingenanto, ngoko uguqulelo lwe IsiJamani lunamandla kukungasemva kwesandi, igumbi lengoma, kunye noshicilelo lwefowuni-umgangatho. Ukucima okunzima okanye abathethi abaninzi abathe bathelekiswa baya kubangela ukuba uqinisekiso lubi.Ukuba ushicilelo luya buyela ngaphandle kokusebenziswa, thumela i-imeyili contact@free.ai ngefayili - siya kubuyisela i-token kwaye sijonge ukuba i-engine eyahlukileyo iphatha isandi sakho kakuhle.